Breaking Three Cords

The architecture for university IP management is deeply entrenched. It is held in place by a set of three narratives, each of which is readily challenged, but together have such a satisfying outcome that it is difficult for administrators not to like them.   In a sense, administrators have been served a big bowl of mind candy for the past thirty years, and they have found themselves unable to defer their taking of as much as they can reach. It has indeed been a very clever scheme, a great practical–but serious, and actually impractical–joke perpetuated on university administrators, by which they have come to give up the values that characterize university inquiry and instruction and adopt in their place values of corporate control and investment.

The Linear Model. One of the key narratives in the set is the linear model of innovation, which starts from basic research at universities, moves to applied research at government and industry, and then to product development, led by investors, from which all future economic public good arises.  The linear model narrative is great for university folks.  It feels good to tell a narrative in which everything depends on university research.  That’s great for funding, of course, but also it’s great just to tell a story in which one’s work is important.  It’s like imagining hitting the winning home run, night after night.  It’s a good feeling and it probably contributes mental health benefits as well.  Especially if one really does hit winning home runs night after night. However, if one doesn’t hit homers, or much of anything at all, the next best thing is to fix one’s belief that one is hitting homers all the time, or just about to do so.  As Peirce argued, belief is more enjoyable than doubt, and there is something satisfying in latching onto those things that appeal to one’s already fixed beliefs.

Investors also like this narrative, as it creates a critical place for capital in the process.  Government funding should carry ideas aloft, and at some point release them so that they glide down through applied validation until they are sufficiently de-risked so that the powerful engines of private investment can be lit and new products change the world.  What’s not to like if one is an investor.  It is a truly happy story, in which investors are pivotal.  No investment, no innovation.  Sad, but avoidable.

There are other narratives of innovation, of course.  There is the narrative that Vannevar Bush sets out in Modern Arms and Free Men that tells of how innovation took place during World War II, led by “gadgeteers” who learned their technology from amateur radio and working on cars, mixed with ace scientists from industry and universities pushed up against a military that needed more than what it thought it needed more of.  And there is the narratives that Steven Johnson describes in Where Good Ideas Come From, including the wildly successful “networked, non-market” approach that dominates the list of breakthrough, transformative ideas of the past century.  And there is the argument that Matt Ridley makes in The Rational Optimist, that science lagged the industrial revolution, and that tinkerers spurred innovation, and fed science stuff that it could later turn into theories and explanations.

In a similar way, companies like Apple, Microsoft, and Facebook developed new technologies that create a need for all sorts of science, or at least computer science, but themselves drew from all sorts of things–including basic research done in years past at universities, but also in industry, and also stuff they did on their own.  Even in the history of warfarin we find that a kind of applied research came first, then a rather unsuccessful application (clinical), then a successful application (rat poison), then a switch of the successful application to the previously less successful application (clinical, again, with a different compound).  The research never did solve the original question of how to protect cattle from eating moldy sweet clover hay, and the university science of how the compound acted was developed thirty years after successful clinical use.

Consider Edison and Tesla–they not only worked in applied areas but also explored the unknown, doing very basic researchy things, just not under the auspices of a university. A hundred years ago, the debate in the big research labs these kinds of inventor-showman–leaders created was whether you let scientists muddle until they have something, and then try to make it, or try to make something and where the engineers and craftsmen cannot get it to work, hand the problem over to scientists.

We live in a world in which the narrative “science first” is resounding, at least from university administrators.  But there to doubt the strength of the claim.  That is, there is no real question that scientific discoveries, like other discoveries, do enable, eventually, useful stuff to get done.  There always is a way to trace back from any useful item to natural history that enables it, and to ascribe the findings of natural history to science.  Put another way, there appear to be a variety of narratives that operate to point out how innovation comes about, and some focus on the field work rather than theory, or on the prepared mind combined with chance, or on a lot of capable folks with very different experiences coming together to work on problems.

Todorov advises us that in history there is no primitive narrative, no single chain of events that when all facts and evidence are assembled compel a single necessary and sufficient account.  We do not live in a world of primitive narratives, but one in which primitive narratives stand for things.  Primitive narratives are themselves figures of thought, a kind of fiction that engages the world and frames it.  At best a rule of thumb, at worst a delusion that prevents doubt that might lead to epiphany.  The world does not operate as primitive narratives, but there is a social, if not political and financial attraction, to having a single narrative that fixes belief and sticking to it.  Proposing other narratives is not a battle for control of the One narrative, but that is often what it is taken to mean.  Rather, and especially for innovation, we benefit from an exploration of the space, a mapping of possibilities.

The thing about innovation is, like colored swans, it shows up where it’s not expected, not part of the consensus, not the thing that is efficient, established, credentialed, planned, or even legal.  Innovation defies management until management finds a way to benefit from it.  The problem for leaders of the established order is that they have a difficult time raising the prospect of any innovation that would render them not the leaders of the established order.  Management likes incremental improvement, mitigation of faults, and execution of the next step on a roadmap, with delight when something turns out to be even better than planned.  Thus, any competing narrative of innovation, were it not to confirm the established order, or not play to the strengths of those leading the established order must be suppressed, ridiculed, or co-opted into another, and acceptable, form.

Bayh-Dole as a Vesting Statute. A second key narrative is that of Bayh-Dole.  In this narrative, which is a misconstruction of the actual law, the government has done a sucky job of managing inventions made with government support, and so must turn to universities of all things, to get it right.  In this narrative, Bayh-Dole transfers ownership of federally supported inventions to university administrators and tasks them with finding commercialization partners who will agree to turn inventions into products for profit, sharing that profit with the universities.  In this narrative, universities already in possession of the critical scientific effort to lead basic research are now also given the mandate to turn that research into gold.  That is, the transition from basic research to applied research is to be done through patent licensing.  For each invention derived from basic research, a university will select an exclusive licensee to invest the private resources necessary to create new products. The creation of these new products will confirm the wisdom of providing tens of billions of dollars a year in federal support to university research.

To put it more precisely, the university’s show of diligent effort in trying to find such licensees is sufficient to confirm the wisdom of using universities to conduct basic research: if licensees are not forthcoming, that is a problem in industry, in the marketplace for inventions, in the innovative capacity of a region, suggesting that the government should intervene, there, as well and provide the investment capital where private investors choose not to invest, and to start the companies where private companies choose not to adopt and operate, and to subsidize industry so that it has incentives to take the license deals on offer from universities.

Bayh-Dole’s role in this narrative is to provide a federal policy mandate to universities to adopt a monopoly exclusive commercial licensing model for faculty research results. The idea is that federal research support is the government’s gift to universities to advance institutional finances through patent license income.  The genius of Bayh-Dole is that it confirms as federal policy that institutional money-seeking through patent positions is in the national, and public interest.  In this narrative, the interests of the faculty investigator, and of the faculty or staff or volunteer or student inventors, the broad dynamics of collaboration in research, and the potential for collaboration in invention management all are made subordinate to an institutional mandate to own, to control, and to license on the condition of product development and payment. Federal law requires faculty and others to conform their expectations to what the institutional model offers: a share of royalties in exchange for helping to make the licensing program “successful”–that is, profitable for the institution.

This then is the “entrepreneurial” university, by which is meant the “profit-seeking” university, using a broad claim of ownership to research scholarship as the foundation.  One might say, this is the “entrepreneurial” administrative university, not the “entrepreneurial” faculty or creative class university.  That is, in the administrative version of Bayh-Dole, the law cuts in administrators as the key catalysts by which basic research becomes applied. Without broad ownership claims, enacted absolutely, early, and often, the invention won’t be used, developed, or turned into product.  It is up to administrators to see that inventions are owned, patented, licensed, and “commercialized.” That is the narrative, anyway.

One might also say that this is a university that as an institution takes research result ownership positions that exclude others (even those that created the results) and in so doing seeks to create “value” in those positions arising from what others are denied, so that this value is recognized by a licensee willing to share the value of that exclusion:

• “How much is it worth to you for us to keep this invention from your competitors?”
• “How much is it worth to you for us to limit the circulation of this invention in preference to your products and time lines for future development?”
• “How much is it worth to you for us to prevent others from using this invention in its present form so that they have to purchase a product from you years from now?”

These are the fundamental questions of the “entrepreneurial” administrative university. They don’t figure, generally, in the administrative narrative, but they are the operative questions in fact at the point of practice.

Wild Success. A third narrative has to do with the effectiveness of universities in carrying out this mandate of developing new knowledge through basic research, taking on a national mandate to own and license rather than support and publish inventions.  Here, university administrators are ready to report “successes”–the few instances in which an invention is licensed for profit, or at least licensed with the expectation of profit, or at least the potential for the expectation of profit.  The successes are meant as metonymies–as instances that serve to stand for the expectation that a lot of inventions will be successful this way, that there are other successful inventions that no one has room to describe in one publication, that there will be more of these in the future, that the future is ripe with potential for impact.  In all, it is a very positive position:  university research is and will continue to be the engine of the economy, and the core of that engine is the institutional patent, offered for exclusive license, the instrument deemed to be the most important for attracting investment to create commercial product.

The invention metrics literature is strewn with efforts to characterize the impacts of university research and the impacts of university licensing.  recent set of articles proposed 56 “indicators of impact” for research, including “number of publications” and “meetings with important people” and “esteem surveys”.  Not one of the proposed indicators, though, was of the form of “innovation” or “new product” or “jobs created” or “new discoveries inspired.”  Similarly with the AUTM licensing survey, which reports aggregate numbers across categories distributed in time across years.  Thus, “research expenditures” by a university is associated (by proximity in a table) with “invention disclosures,” “patent applications,” “patents issued,” “licenses,” and “income.”  These metrics are distributed across perhaps a decade of time, from the time money is spent to the time an invention report is filed, a patent application is filed, a patent (perhaps) issues [we are three to five years downstream already], to a license and income [another three to five years, from license to income based on sales or realized equity].  Yet academics and patent administrators alike have been known to divide one into another to show “efficiency”–as if the money spent last year by the university has anything to do with a patent issuing this year, which was the result of an invention made in the spending at the university four or five years ago.

Even then, there’s no reason to consider the aggregate spending as a measure of anything other than scale.  If a bunch of that spending is on clinical trials, then there’s no university IP in the game from that spending.  It doesn’t mean a thing, since universities generally don’t take IP positions in clinical trials work unless it is for a protocol they are proposing for a compound they own, in which case they shouldn’t be doing the trial in the first place.  If one wanted to consider the amount of money spent in a given grant project relative to the number of patentable inventions it produces over time, that might be an interesting number if the purpose of the research is to produce patentable inventions.  If not, again, it is meaningless, really, as an indicator of productivity or efficiency.  May was well count how many times a baseball team burps during a game.  Great if your business is in marketing flatulence aids.  Meaningless in terms of how baseball is played and games are won.  A university patent on a university research invention is, in most respects, a flatulence aid.

The upshot of the AUTM licensing survey statistics is that they present a disconnected set of aggregate data that collectively chart the overall demands on an office in a year, but do not have much by way of relationship to one another.  The invention reports received in a year have only somewhat to do with the patent applications filed in that year (and also with the patent applications not filed or abandoned), and the patents issuing in a year have nothing to do, really, with the applications filed.  And all of the filing activity depends in large part on having a budget to spend on patenting–either from royalty income, a university budget, or from licensees willing to pay for the university to service the patenting activity.  Yet no one reports the source of funds for patenting.  One could easily argue that having unlimited funds for patenting could create a huge disaster for patent management–spending on patents that will never come into play, or spending to justify claims to inventions that should never be held by a university, or held as proprietary, or pushed into a monopoly preferred licensing model.  More patents may mean more waste, more noise, more overhead, more distraction.  It may not mean anything at all with regard to research “efficiency” or “productivity”–certainly nothing about research impact.

The AUTM licensing survey, furthermore, does not report anything specific to Bayh-Dole, that pivotal element of the second narrative so loved by university administrators.  AUTM does not report on subject inventions, does not report the status of those inventions, does not report the date of first commercial use or sale for those inventions.  And these are metrics that the standard patent rights clause expects universities to report, if the funding agencies get around to asking for utilization reports. Universities ought to be reporting this information annually anyway.   Except they don’t.  And Bayh-Dole makes all such reporting exempt from FOIA.

Here then is another change to Bayh-Dole that would be welcome.  Under 35 USC 202(c)(5), Bayh-Dole requires agencies to treat any periodic reporting on utilization of subject inventions to be treated as “privileged and confidential.”  Modify Bayh-Dole to require annual reporting by each university or nonprofit contractor, and require that for each subject invention reported, that the university report for public release (1) the date on which it elects whether or not to retain title, and whether it so elects to retain title; (2) the status of patenting, whether a provisional or utility application has been filed, or a patent has issued, and if there are foreign applications or the prospect for continuations;  (3) the status of the invention, whether licensed or not (and if licensed, whether exclusive), whether being further developed or not (and if so, whether by the university or others), whether being used or not (by the university, or by others, or both); and (4) the date of first commercial sale or use (which is specified in 37 CFR 401.14(a)(h)).

Such information can be disclosed publicly without revealing the nature of the invention, the specifics of the patent work, the identity of any licensee, or any financial or commercial considerations.  This information would show, for a given set of reported inventions, the core institutional positions with regard to those inventions–ownership, development, licensing, commercial sale or use.  Doing so would provide context for the condition of inventions brought into a university invention management program, whether by compulsion or voluntarily.  If a university is taking on more work than it can service, there’s a problem.  But one would never know it from the reporting of aggregate figures and success stories, which hide a few out of scale transactions in 20 years of lump sums.

If one wanted to really get at the cost and return of any such university patent licensing program, the university would also report (5) the total expenditures and obligations directed at the licensing operation, including salaries, operations, legal expenses, payments to inventors and others, professional development, leased space, and an estimate of associated indirect costs; and (6) total income from licensing.  The NSF, perhaps, ought to include these figures in its survey of university R&D expenditures.  These numbers in any given year may be out of sync–one could spend this year, and appear to lose money, but as a result have a big license with lots of income for the next decade.  Over a period of time, however, it should be possible to see whether a program is consistently running at breakeven or better.  With a “big hit” license, likely it will, and for some time.  It may be that even without a “big hit” such programs should be subsidized–there are good arguments for that, especially as a program is directed at building standards and broad access rather than at monopoly positions.  But one would never know how to make such an argument without seeing the actual financials.

Unfortunately, it appears that the primary rationale administrators have for subsidizing a patent licensing program is the hunt for the “big hit” deal.  The purpose of such a subsidy is to be able to file more patent applications–and thereby claim a larger share of faculty research–without having any sponsor, donor, or licensee willing to pick up some or all of the costs.  The operative idea is that it is worth the institution’s money to prevent anything from being successful without the university having a stake in the success, and more than that, without the university controlling the choice, structure, and its stake in the transaction.  Under this approach, the subsidy is to expand the capacity to withhold (from inventors, from other investigators, from practitioners, from industry, from entrepreneurs, from investors), but not necessarily to promote the use of the invention.

Administrators are easily sold on this kind of subsidy as it is described as an “investment” or “entrepreneurial.”  However, they are not willing, like recalcitrant gamblers, to present the outlays and the returns.  There is always the hope of a hundred million dollar invention, like a lottery ticket or a bay horse, that will wash away all losses and turn the “investment” into “success”.  Again, all it takes is one invention every twenty years to go this way, providing slush money for the past expenditures.  The foolishness in this approach is in thinking that the hundred million dollar invention is somehow more likely if the university takes all inventions from faculty, to ensure that none escape “out the back door.” An alternate suggestion is that the hundred million dollar invention comes about as a result of outlier, black swan luck, and the best chance for a university to participate in any financial upside is goodwill and visible support.

University administrators, however, do not report the metrics that would allow anyone to assess whether their particular approaches are “working.”  While many quantitative figures are reported, none of the ones expected by standard patent rights clause under Bayh-Dole are reported.  None of the actual operating expenses are reported.  No status report is given so one can see the state of the portfolio, to see whether the five patent applications this year are continuations of the five patent applications last year, or divisionals, or continuations in part, or just refiling a provisional application as a utility application.  Those sorts of figures–real figures, useful figures–are not reported.

What we have, then, in the United States is this tri-fold narrative comprised of the linear model that places basic research first and in the hands of the universities; a misconstruction of the Bayh-Dole Act, which provides a mandate for institutional ownership, patenting, and “commercialization” licensing; and a free-standing repeated but unverifiable claims to “success” or the the “potential” for that success, without providing the backing figures that would allow an assessment of the claim.  Together, these narratives create a tri-fold architecture that is not easily challenged.  So long as people–academics, policy-makers, administrators, faculty, companies–believe that basic research comes first in all narratives of innovation, that the Bayh-Dole Act authorizes (or should have authorized but for the nasty ignorant uncooperative Supreme Court) university ownership of inventions, and that the result of the research and licensing is “success” as evidenced by a plethora of upward trending quantitative data, then the complex is held in place.  Challenge any one of these, and the other two rush in to the defense.

If Bayh-Dole does not mandate institutional compulsory ownership of inventions, then it should or the whole technology transfer system–the wildly successful technology transfer system–is at risk, as if that system relies on institutional compulsion.  If the wild success of the system is challenged, then federal mandate comes to the defense:  the government was failing to “commercialize” and this system has the mandate of the government, of the people.  How can anyone look only at the money, or at the unlicensed patents?  That’s sour grapes and ignorance.  Look at the licensing metrics.  Look on the bright side.

If one challenges the linear model, that is taken as a direct assault on universities as recipients of federal funding for research.  The institutions are optimized for just the symbiotic relationship with the government, with the public, with the states (in the case of public universities), with industry, and with the students.  There is a good front of rhetoric facing each direction.  Yet at the heart of it, none of the narratives is holding up.  At the heart of it, things are not pretty, but for the polish of the infographics and the authoritative look of the published statistics:  the public domain is being been throttled, personal initiative and insights of investigators co-opted, academic freedom and independence of investigation suppressed, collaboration with industry disrupted, inventors disrespected and stripped of rights, federal regulations and a pivotal Supreme Court ruling ignored and misrepresented, opportunities destroyed and lost, monopolies without redeeming consequence created, direct practice of scholarship disrupted, publication and instruction undermined, and forms of support abandoned in favor of reliance on venture investors.  It is not a pretty picture, other than the rosy reporting of the narratives.

Of all the outcomes of this three-cord narrative–linear model, university ownership, wild success–the saddest perhaps is the loss of administrators tuned to the values of independent faculty research.  Those values appear to be lost–if not utterly forgotten, so suppressed that no one is willing to raise them in discussion, let alone in policy, as the real framework on which university involvement in inventions is based.  Bayh-Dole provided an opportunity and a challenge.  The opportunity was to bring more of what was federally supported into the ambit of invention management resources already available to faculty inventors. The challenge was to use the patent system in ways consistent with university faculty independence to promote the use of inventions, free competition among industry players, help for the small companies, and lower administrative costs for all.  Instead we have got something very different.  We have all federally supported inventions inserted into the system until it has become bloated, distorted, re-rationalized; we have the eradication of external and multiple avenues by which inventions might be made available, or managed; and we have the substitution of the chase for a few big licensing transactions instead of multiple daily efforts to engage creative people–scientific, technical, financial, and cultural–with the ideas and possibilities and capabilities of university hosted talent.

The talent and diligent and good intentions of individuals is not the issue.  The perverse, and pervasive, administrative architecture of three flawed, limited narrative cords is.